Zhang (2006) 發現公司價值資訊的不確定性會使市場對好壞資訊的反應不足造成股票報酬率差異的產生,亦即資訊的不確定性高的股票,投資者對好壞資訊的反應比資訊的不確定性低股票較不足,股票報酬率也相對較高或較低,因此造成可能的套利空間。 在不確定性代理變數由1. 公司規模2. 公司年齡3. 股票報酬波動性4. 分析師的報導遮蓋率5. 分析師預測的分散度6.現金流量的波動個別討論之外並加入以公司的財務盈餘預測的準確度高低來驗證:得之如果公司過去盈餘預測準確程度高(即預測EPS的標準差低),則分析師預測修正時,股票收益率比過去盈餘預測準確性低的公司(即預測EPS的標準差高)為低;相反的,如果公司過去盈餘預測準確程度低(即預測EPS的標準差高),則分析師預測修正時,股票收益率比過去盈餘預測準確性高的公司(即預測EPS的標準差低)為高。惟只有上市與上櫃公司樣本合併之下才支持此假說。
Zhang (2006) discovered the uncertain information of firm value would cause market investors’ insufficient response toward good or bad news. Which makes the stock returns difference, i.e. those stocks possess higher information uncertainty, investor will have under-reactions followed the good or bad news than those with low information uncertainty’s, their stock return will relatively higher or lower. Hence, there would be certain opportunity for interest arbitration. I use the 6 uncertainty proxies: Firm Size, Firm Age, Stock Volatility, Analyst Forecast Coverage,. Analyst Forecast Dispersion,. Cash Flow Volatility chosen by Zhang plus Corporate Forecast Accuracy to discuss stock returns changes via analyst forecast revise, and find out that if the firms with higher corporate forecast accuracy eventually will reverse the degree of information uncertainty (negative correlate), therefore, the future stock returns should be lower upon the analyst forecast revise. And if the firms with lower corporate forecast accuracy (higher EPS standard deviation), the future stock returns should be higher. This is empirically tested , but only be supported when the samples of TWSE & OTC are combined together.